https://orcid.org/0000-0002-9833-4425
Figures
Abstract
The world-wide prevalence of myopia (nearsightedness) is increasing, but its pathogenesis is incompletely understood. Among many putative mechanisms, laboratory and clinical findings have implicated circadian biology in the etiology of myopia. Consistent with a circadian hypothesis, we recently reported a marked variability in diurnal patterns of gene expression in two crucial tissues controlling post-natal refractive development ‐ the retina and choroid–at the onset of form-deprivation myopia in chick, a widely studied and validated model. To extend these observations, we assayed gene expression by RNA-Seq in retina and choroid during the progression of established unilateral form-deprivation myopia of chick. We assayed gene expression every 4 hours during a single day from myopic and contralateral control eyes. Retinal and choroidal gene expression in myopic vs. control eyes during myopia progression differed strikingly at discrete times during the day. Very few differentially expressed genes occurred at more than one time in either tissue during progressing myopia. Similarly, Gene Set Enrichment Analysis pathways varied markedly by time during the day. Some of the differentially expressed genes in progressing myopia coincided with candidate genes for human myopia, but only partially corresponded with genes previously identified at myopia onset. Considering other laboratory findings and human genetics and epidemiology, these results further link circadian biology to the pathogenesis of myopia; but they also point to important mechanistic differences between the onset of myopia and the progression of established myopia. Future laboratory and clinical investigations should systematically incorporate circadian mechanisms in studying the etiology of myopia and in seeking more effective treatments to normalize eye growth in children.
Citation: Stone RA, Tobias JW, Wei W, Carlstedt X, Zhang L, Iuvone PM, et al. (2024) Diurnal gene expression patterns in retina and choroid distinguish myopia progression from myopia onset. PLoS ONE 19(7): e0307091. https://doi.org/10.1371/journal.pone.0307091
Editor: Steven Barnes, Doheny Eye Institute/UCLA, UNITED STATES OF AMERICA
Received: April 10, 2024; Accepted: June 30, 2024; Published: July 19, 2024
Copyright: © 2024 Stone et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The data generated in this study were deposited in GEO (Gene Expression Omnibus) with accession number GSE261232 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE261232).
Funding: Supported by National Institutes of Health grants R01 EY025307 (DLN), R01 EY016435 (PMI), P30 EY001583 (RAS), P30 EY006360 (PMI), an unrestricted grant from Research to Prevent Blindness (RAS), a Challenge Grant to Emory Eye Center from Research to Prevent Blindness (PMI), and the Paul and Evanina Bell Mackall Foundation Trust (RAS). The funders did not play any role in the study design, data collection and analysis, decision to publish or preparation of the manuscript.
Competing interests: The authors have read the journal’s policy and have the following competing interests: RAS serves as an advisor to iView Therapeutics, Inc. (https://www.iviewtherapeutics.com/). PMI serves as an advisor to Spave Science, Inc. (https://www.spavescience.com/) and to NeuroRays, LLC (https://neurorays.com). This does not alter our adherence to PLOS ONE policies on sharing data and materials. There are no patents, products in development or marketed products associated with this research to declare.
Introduction
The coupling of the optics in the front of the eye with its axial length determines ocular refraction. During childhood, an active regulatory process synchronizes the expanding eye length with the corneal and lens powers; malfunctioning of this process yields refractive errors. In myopia (nearsightedness), the most common refractive error, the eye is relatively long for the optical power of the cornea and lens; distant images are blurred because their focus lies in front of the retinal photoreceptors [1]. While varying between countries, the prevalence of myopia is increasing markedly world-wide. It is reaching a prevalence in young adults of some 40–50% in the United States and Europe and 80–90% in regions of East and Southeast Asia [2–4]. By 2050, estimates suggest that some 50% of the world’s population will be myopic [5]. Besides blurred vision, myopia increases the risks for various retinal detachments, macula and retinal degenerations, glaucoma and certain forms of cataract; thus, myopia contributes to many blinding diseases of adulthood [6]. Notwithstanding many decades of clinical and basic research, many hypotheses and much speculation, the etiologic mechanism(s) for the altered eye development underlying myopia or for its increasing prevalence are unclear.
Convincing evidence, initially from animals but confirmed in children, has linked refractive development and myopia to visual input; and the controlling mechanism(s) localizes largely to the retina [7–10]. Not only in chicks but also in young mammals, for example, ipsilateral form-deprivation myopia follows the wearing of an image-degrading diffuser over an eye [11]. Visual input also influences the thickness of the choroid, the tissue adjacent to the retina. Interactions of the choroid with both retina and sclera are postulated to occur within a retina-to-choroid-to-sclera signaling cascade controlling overall ocular growth and refraction [12–14]. A vast number of signaling molecules, transcription factors, enzymes and biological pathways have been associated with myopia, but no coherent framework has yet emerged that clearly provides an explanation for the pathogenesis of clinical myopia [9].
Numerous laboratory and clinical findings have recently implicated circadian disruption in the development of myopia [9, 15–19]. Dimensions of both animal and human eyes fluctuate during the day, and these fluctuations seemingly affect overall eye growth and refractive development [13]. Experimental myopias of chick and mouse alter the retinal expression of clock and circadian rhythm-related genes [20–24]. In mice, Bmal1 clock gene knockout in retina produces myopia [25]. Melanopsin gene knockout in the retina of mice modifies normal eye development and increases experimental myopia [26]. Also in mice, obliterating intrinsically photosensitive retinal ganglion cells (ipRGCs) that project to the hypothalamus to regulate circadian rhythms curbs myopia [26, 27]. In young adult humans, short-term effects of light stimulation of ipRGCs suggest an interaction of ipRGC signaling with myopia stimulation [28]. Furthermore, the defocus and blurring of visual images that induce refractive errors in chick each modify the diurnal pattern of clock and circadian rhythm-related gene expression [29]. In human GWAS studies, the hundreds of specific genes and genetic loci associated with myopia and/or refractive error include genes mapping to genetic networks involving circadian control and light sensitivity [16, 19].
Because of these findings implicating circadian biology and because of diurnal changes in gene expression in many tissues [30, 31], we previously assayed gene expression in retina and choroid at six times over a day in chicks at the onset of unilateral form-deprivation myopia induced by unilateral diffuser wear [18], a well-established myopia model [32, 33]. We chose these two tissues because the visual signals regulating refractive development presumably originate in the retina and act at or through the choroid to modulate scleral growth and hence ocular size and refraction [11, 13, 14, 33, 34]. In both retina and choroid, we found surprisingly high variability of gene expression differences across the day comparing form-deprived to contralateral control eyes. Both in animals and in humans, the mechanisms acting at the onset of myopia likely differ from those during the progression of established myopia [35, 36]. Thus, we expanded our prior study [18]; but instead of assaying gene expression in retina and choroid during the first day of myopia initiation, we induced unilateral form-deprivation myopia in the same strain of chick and then assayed gene expression on the fifth day of continuous form-deprivation when myopia is both well-established and progressing rapidly [37, 38]. Not only did we find marked variability in gene expression in retina and in choroid over the course of a day during progressing myopia, but we also found that the altered diurnal gene expression during myopia progression differs markedly from that occurring at myopia onset.
Materials and methods
Animals and tissue harvesting
Except for a longer period of myopia duration, we followed the protocol we previously used to assess gene expression changes at myopia onset [18], studying unilateral form deprivation myopia in chicks (Gallus gallus domesticus) of the Cornell-K strain (a closed flock, random-bred for over 60 years). Newly hatched Chicks (n = 36 chicks) were reared for 12 days under a 12-hour light/12-hour dark cycle with illuminance of ∼300 lux in the cage (Phillips MAS LEDtube HF, 6500K; https://www.lighting.philips.com/main/prof/led-lamps-and-tubes/led-tubes/master-ledtube-instantfit-hf-t8/929001284202_EU/product). Then at ZT 0 (ZT, zeitgeber time; defined as lights on at ZT 0), a plastic image-degrading diffuser attached to a Velcro ring was fastened over the right eye using a complementary Velcro ring glued to the feathers around the eye. Diffusers were worn for four full 24-hour cycles, with the eye beneath the diffuser as “occluded” and the contralateral control eye with intact vision as “open.” Starting on day 5 of diffuser wear, chicks were killed by decapitation without anesthesia in timed groups, with chicks randomly assigned to time. Tissues were acquired at approximately ZT 0, 4, 8, 12, 16, or 20 hours (n = 6 chicks/time/condition), using a previously described tissue sampling procedure [29]. The duration of diffuser-wear corresponded to a time of established, rapidly advancing form-deprivation myopia in this strain of chick [37, 38]. For dissections during the dark phase, chicks were killed under a dim dark yellow light from a photographic safe light (Premier Model SL1012, Doran Manufacturing, Cincinnati OH, USA; ∼0.5 lux). After enucleation, eyes were put onto a bed of ice, hemisected at the ora serrata, and the vitreous was removed. Eyecups were put into Dulbecco’s Modified Eagle’s Medium under sterile and RNAse-free conditions, and the pecten was cut out using a razor blade. Using an oblique dissector/elevator (oblique ends 2mmx10mm, 6 ½”; item #160–720; George Tiemann & Co., Smithtown, NY), the entire extent of both the retina and choroid were dissected off of the sclera, and every attempt was made to remove any adhering RPE from both tissues. However, we cannot exclude the possibility that a small amount of RPE or RNA from RPE remained in some cases. To minimize RNA degradation, RNAase AWAY (ThermoFisher, Waltham, MA) was used to wipe the end of the periosteal elevator between the tissues. The tissues were immediately snap-frozen in liquid nitrogen, stored at −80°C until shipped on dry ice to the University of Pennsylvania. They were then kept at −80°C until processed further. The Institutional Animal Care and Use Committee (IACUC) of the New England College of Optometry approved the research (Approval #A4070-01). The research adhered to the ARVO Statement on the Use of Animals in Ophthalmic and Vision Research, was conducted in compliance with the relevant guidelines and regulations, and is reported in accordance with the ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines.
Library preparation and sequencing
The Next-Generation Sequencing Laboratory in the Penn Genomics and Sequencing Core (RRID:SCR_022383) extracted RNA from retina and choroid tissues with the QIAGEN microRNAeasy kit (QIAGEN, Germantown, MD, USA). RNA was stored at −80°C until shipped frozen to Azenta Life Sciences (South Plainfield, NJ, USA) for further processing and sequencing. Azenta had access only to the Penn-assigned integer labels of the samples but not to the experimental conditions of each sample.
Library preparation and sequencing were conducted by Azenta with PolyA selection and Illumina Sequencing. RNA samples were quantified using Qubit 2.0 Fluorometer (Life Technologies, Carlsbad, CA, USA). RNA integrity was checked using Agilent TapeStation 4200 (Agilent Technologies, Palo Alto, CA, USA). RIN (RNA Integrity Number) values were >9.0 for each sample, with mean RIN = 9.4±0.2 (mean±SD); see S1 Table for values of individual samples. Strand-specific RNA sequencing libraries were prepared by using NEBNext Ultra II Directional RNA Library Prep Kit for Illumina with 1 μg RNA/sample and following the manufacturer’s instructions (NEB, Ipswich, MA, USA). Briefly, the enriched RNAs were fragmented for 8 minutes at 94°C. First strand and second strand cDNA were subsequently synthesized. The second strand of cDNA was marked by incorporating dUTP during the synthesis. cDNA fragments were adenylated at 3’ ends, and indexed adapter was ligated to cDNA fragments. Limited cycle PCR was used for library enrichment. The incorporated dUTP in second strand cDNA quenched the amplification of second strand, which helped to preserve the strand specificity. The sequencing library was validated on the Agilent TapeStation (Agilent Technologies, Palo Alto, CA, USA), and quantified by using Qubit 2.0 Fluorometer (ThermoFisher Scientific, Waltham, MA, USA) as well as by quantitative PCR (KAPA Biosystems, Wilmington, MA, USA). The average insert size was approximately ∼560bp, as measured by Tapestation, with a mean molarity of 71.7nM/insert. The sequencing libraries were clustered on the flowcell. After clustering, the flowcell was loaded on the Illumina instrument (4000 or equivalent) according to manufacturer’s instructions. The samples were sequenced using a 2x150bp Paired End (PE) configuration. Image analysis and base calling were conducted by the Control software. Raw sequence data (.bcl files) generated the sequencer were converted into fastq files and de-multiplexed using Illumina’s bcl2fastq 2.20 software. One mismatch was allowed for index sequence identification. A total of 5.4 billion reads were mapped successfully to the chicken transcriptome, with a mean of 37.7 million mapped reads/sample and with a Mean Quality Score of (35.73±0.07)/sample (mean±SD; S1 Table).
Data analysis
Since the nature of the data is similar, we followed the analysis technique we previously applied to our study of gene expression at myopia onset [18]. With Salmon [39], reads were mapped against the transcriptome described in Ensembl version 105.6 built on the chicken genome assembly GRCg6a. Applying several Bioconductor packages in R [40], transcriptome count data were annotated and summarized to the gene level with tximeta [41] and further annotated with biomaRt [42]. DESeq2 [43] was used for normalizations and statistical analyses.
DESeq2 was used to calculate normalized counts, variance stabilized counts and statistical analyses. For time-course analyses, a reduced model prioritized genes with unequal expression or differential expression across all time points. Included in the computations were baseMeans (the average normalized counts across all samples), p-values, and p-adj values [the p-values corrected for the false discovery rate (FDR) using the Benjamini-Hochberg method].
Based on the expression level of the HINTW (histidine triad nucleotide binding protein W; ENSGALG00000035998) gene, coded on the chicken female W chromosome, this study contained 15 female and 21 male chicks, with 2 or 3 female chicks and 3 or 4 male chicks at each individual time. Sex was removed as a source of variation when examining the factors of primary interest by including sex in the statistical model.
The main outcome analyses included: 1) Gene expressions in each condition (tissue, occluded eye, or open eye) were evaluated independently for changes over time, separate from changes in the other conditions. 2) Following an established analytical approach to unilateral experimental myopia, gene expression levels in occluded eyes were compared to those of its contralateral open control eyes at each time, as the log2-transformed ratio of the normalized means of [(occluded eyes) vs. (contralateral open eyes)] with p-adj values for statistical significance at each time. 3) To identify genes with varying expression over time in occluded eyes different from that in contralateral open eyes, we statistically modeled the interaction of treatment (occluded vs. open eye) with time for each tissue (i.e., occlVopen*time interaction). Differentially expressed genes occurring at more than one time during the day were found with Venn diagrams (http://www.interactivenn.net/index.html) [44].
Using DESeq2 statistics [43], likelihood ratio tests (LRT) were calculated with a reduced model to prioritize genes exhibiting unequal expression across all time points. Using the degPatterns function from the DEGreport package [45], significantly changed genes across the day were clustered into groups demonstrating similar expression patterns; the clusters were displayed in graphs (Figs 1 and 2). The degPatterns function assigned directly the group names and orders in these graphs [45]. To identify pathway enrichments, Gene Set Enrichment Analysis (GSEA; v4.3.2) [46] evaluated pairwise comparisons between occluded and contralateral open eyes; GSEA uses the totality of genes measured to determine whether an a priori statistical ranking of those genes shows a positive or negative bias for a set of genes that corresponds to a pathway or functional group. The test is repeated for all gene sets in a collection. GSEA was performed in pre-ranked mode with the DESeq2 statistic as the ranking metric, and tested against the Canonical Pathways collection (C2:CP) of the Molecular Signatures Database (MSigDB; v2023.2.Hs; https://www.gsea-msigdb.org/gsea/msigdb). Based on preliminary comparisons of the gene sets in the Human Molecular Signatures Database, we selected the C2:CP curated collection because its pathway classifications supplied useful pathways (e.g., structure, inflammation, neurotransmission, peptide signaling, photoreception, etc.) to generate hypotheses for myopia mechanisms.
The clustered patterns of variable gene expression during the day are shown for the retina: occluded eyes (A) and contralateral open eyes (B); and for the choroid: occluded eyes (C) and contralateral open eyes (D). Subsets of the varying genes generated these expression patterns using statistical criteria described in the text, with the number of varying genes used to form each group above each panel. The densities of the tracings in each group reflect the number of genes conforming to each specific cluster. To illustrate the light:dark transitions, the bottom bars on each panel indicate the light phase (white bar) beginning at ZT0, and the dark phase (black bar) beginning at ZT12 and ending at ZT0. See S2A–S2D Table for the variability data of all genes over time for each tissue and eye, ranked by p-adj values. Abscissa: times of tissue sampling in ZT (hours). Ordinate or Z-scores: 0 = mean, with non-zero values corresponding to ± S.D.
The groups into which genes clustered for the OcclVopen*time interactions over one day are shown for retina (A) and choroid (B). The limited number of genes meeting the criteria of p-adj<0.05 or p-adj<0.1 necessitated applying weaker statistical criteria to generate these groupings (see Results text). Most but not all interacting genes showed expression patterns demonstrated by these identified groups. The number of genes assignable to the individual groups appear above each panel. To illustrate light:dark transitions, the bottom bars on each panel signify the light phase (white bar) beginning at ZT0, and the dark phase (black bar) beginning at ZT12 and ending at ZT0. See S4A and S4B Table for the complete list of genes in each tissue in the occlVopen*time interaction category, including group assignment, gene names and descriptions, ranked by p-adj values. S1A Fig (retina) and S1B Fig (choroid) show heatmaps for the individual genes in the occlVopen*time interactions category meeting the p-adj<0.01 criterion. Abscissa: times of tissue sampling in ZT (hours). Ordinate or Z-scores: 0 = mean, with non-zero values corresponding ± S.D.
To compare chick genes identified in myopia progression to candidate human genes for myopia, we supplemented a recent summary of human candidate genes linked to myopia [20] with additional, non-overlapping genes from five other recent human genetic studies [47–51]. The chicken transcriptome (as its Ensembl gene identifications) was mapped via orthologs to the human gene symbols. A Chi-square test with Yates’ correction was used to compare the human gene list to the chick genes with statistically significant likelihood ratios in the occluded vs. open eye assessments and with the statistically significant likelihood ratios in the occlVopen*time interactions. Unless indicated otherwise, a value of p-adj or FDR less than 0.05 was judged statistically significant.
Results
The present study on myopia progression parallels the design and analysis of our prior study of myopia onset [18]. For readers interested in comparing the two reports, selected findings at myopia onset are included here. As before [18], tissue was the main feature that differentiated gene expression in the samples. Sex (identified from HINTW gene expression; see Methods) provided a less robust differentiating feature, but sex was included in the statistical models to eliminate it as a contributing variable. The effects of eye and time were assessed in retina and choroid separately. Samples from retina generally displayed less variability than those from choroid.
The variation of gene expression over time
For the ∼17,000 genes identified by Ensembl geneIDs, some 25% of genes in retina varied over the sampling times, and 31–37% varied in choroid, depending on whether the eye was occluded or open (Tables 1 and S2A–S2D). Since only one 24-hour day was studied and since we used neither constant light nor constant dark environments, we cannot distinguish between diurnal or circadian gene expression patterns and cannot separate diurnal/circadian effects from acute light effects [52].
To identify the various patterns of daily gene expression changes for each tissue and eye, we grouped the changes into distinct expression patterns over the day. Because of the large number of varying genes, we reduced this assessment to the most statistically significant subsets of the changing genes. For retina, the cutoff for the clustering models was p-adj = 1xe-10 (690 genes, occluded eyes; 865 genes, open eyes); for choroid, the cutoff was p-adj = 1xe-7 (759 genes, occluded eyes; 757 genes, open eyes). In retina, the variable genes clustered into 11 groups for both occluded and open eyes (Fig 1). In choroid, the variable genes clustered into 9 groups for occluded eyes and into 11 groups for open eyes (Fig 1). From these complex clusters of diurnal gene expression in each tissue and eye, the highest or lowest level of expression of individual genes occurred near the end of the light phase for many but not all gene expression patterns.
Sampling time impacts occluded vs. open eye differences in gene expression
Tissue levels of gene expression, neurotransmitters or proteins are commonly compared between experimental and contralateral control eyes in seeking mechanisms underlying myopia, including form-deprivation myopia [9, 15, 18]. By the statistical criterion of p-adj<0.05, 121 genes in retina were differentially expressed at ZT04; and 341 genes in choroid were differentially expressed at ZT12. For either tissue at the other times, fewer or no genes were differentially expressed (Tables 2 and S3A and S3B). In the retina, the times with fewer gene expression differences occurred mostly in the dark phase or at the dark-to-light transition of ZT00; in the choroid, few gene expression differences occurred either in the light or the dark phases, with the preponderance of altered choroidal genes occurring at the end of the light phase, ZT12.
For retina or choroid, the genes in the “overall” category were identified by considering concurrently all replicates at all time points and prioritizing genes with occluded responses similar in direction and magnitude at all time points. The genes in the overall category were not required to differ significantly between occluded and open eyes at every time point. The overall category included only a small number of genes that differed statistically between occluded and open eyes over the full day in each tissue (Tables 2 and 3 and S3A and S3B).
Occluded vs. open eye differences interacted with time for a limited number of genes
We modeled the interaction of differential gene expression in occluded vs. open eyes with time (“occlVopen*time” category) for retina and for choroid. This analysis identified genes where the expression patterns in occluded eyes and contralateral open eyes differed statistically from each other across the day (Tables 4 and S4A and S4B). In seeking to identify the patterns of the differential gene expressions between the two eyes underlying the occlVopen*time interactions, the numbers of genes meeting either the p-adj<0.05 criterion (retina, n = 37; choroid, n = 7) or the p-adj<0.1 criterion (retina, n = 48; choroid, n = 12) were too few to cluster the gene expression patterns over time into groups. To increase the gene numbers for modeling, we selected a p-adj<0.4 for retina and approximately p-adj<0.797 for choroid that yielded 185 retinal genes and 176 choroidal genes that clustered into patterns where gene expression patterns differed between the two eyes. The genes in the occlVopen*time interaction category identified by these broadened statistical criteria are shown schematically in Fig 2 and are listed in S4A (retina) and S4B (choroid) Table. Depending on the modeled group, specific gene expressions in occluded eyes were higher or lower than the open eyes at some or most times. Occasionally, the relative gene expression levels became inverted between the two eyes over the day (e.g., Fig 2A, group 14). Heatmaps display the diurnal expression patterns of genes with statistically significant occlVopen*time interactions (p-adj<0.1) for retina and choroid (S1A and S1B Fig).
Only a few genes developed occluded vs. open eye expression differences at more than one time
Genes with differential expression levels between occluded vs. open eyes at more than one time were assessed with Venn diagrams (Fig 3) [44]. Regardless of how many genes showed occluded vs. open eye differences at a specific time (Table 2), very few genes exhibited occluded vs. open eye expression differences at more than one time: retina: 7 genes; choroid: 3 genes (Fig 3 and Tables 5 and S5). Most genes decreased at each common time point, although one gene in retina and two genes in choroid altered the direction of change depending on sampling time. Four retinal genes and one choroidal gene were altered at 2 consecutive times. Of the single gene in retina affected at 3 times, only two of the times were consecutive (S5 Table).
The Venn diagrams identify the number of genes with differential expression levels at more than one time comparing occluded vs. open eyes in retina and in choroid, with the criterion of p-adj<0.05 for occluded vs. open eyes (see S3A and S3B Table). The different shapes of the Venn diagrams in the two panels occurred because the number of times with differentially expressed genes (Table 2) differed between retina (5 times) and choroid (4 times). S5 Table indicates times, the number of genes differentially expressed at the indicated times, the specific genes altered at more than one time, the directions of gene expression change and the log2 fold changes for both tissues.
S6A (retina) and S6B (choroid) Table list all gene data comparing occluded vs. open eyes at any p-adj level, with gene names and fold changes.
Biochemical and signaling pathways during myopia progression
To identify signaling, structural and metabolic pathways potentially involved in progressing myopia in retina and choroid, we performed Gene Set Enrichment Analysis (GSEA) [46] on the relative expression of genes in occluded vs. contralateral open eyes (S7A and S7B Table). We found pathways in the GSEA report enriched in the occluded eyes relative to contralateral open eyes that corresponded to genes with increased expression in occluded relative to the contralateral eyes. We also identified pathways enriched in the open eyes relative to contralateral occluded eyes that corresponded to pathways from genes with decreased expression in occluded eyes relative to contralateral open eyes. GSEA uses unfiltered but ranked gene lists to detect pathways from genes moving in the same direction and identified pathways from each tissue and each time condition, even if few or no individual genes under some conditions met the statistical p-adj<0.05 criterion for differential gene expression. Based on an FDR<0.05 for pathway identification, we evaluated GSEA for the occlVopen*time interactions, for the overall category and for each time (see S3A, S3B, S4A, S4B, S7A and S7B Tables).
Some differentially expressed chick genes corresponded to candidate genes for human myopia
We compared candidate human genes from recent publications (see Methods) [20, 47–51] to the differentially expressed genes identified in the progression of form-deprivation myopia in chicks (p-adj<0.05; S3A, S3B, S4A and S4B Tables). Only a limited number of candidate human genes corresponded to the differentially expressed genes in established, progressing chick form-deprivation myopia (Table 6). The retinal genes in chick corresponding to human genes were found mostly at chick light-phase sampling time ZT04, consistent with the preponderance of differentially expressed chick retinal genes at this time (Table 2). However, two overlapping chick/human genes also were identified in retina at sampling time ZT16, in the middle of the dark phase. The choroidal genes in chick corresponding to human genes were found only at sampling time ZT12, also consistent with the preponderance of differentially expressed chick genes in choroid at this time (Table 2). No genes in the occlVopen*time interaction or overall categories corresponded significantly with the human gene list.
Discussion
Consistent with the evidence linking refractive development with circadian biology described in the Introduction, the expression levels of many genes during established and progressing myopia depend on the time of day in both retina and choroid (Table 1 and S2A, S2B, S2C and S2D), with expression patterns varying by gene, tissue, and the eye’s occluded/open status (Fig 1). In terms of the myopia analysis of occluded vs. open eye comparisons, the central finding during progression is that gene expression differences between occluded myopic and contralateral open eyes also depend on tissue sampling time during the day (Table 2 and S3A and S3B). Most differentially expressed retinal genes were identified during the light phase (ZT00-ZT12), and the largest number of gene expression differences between occluded and open eyes during myopia progression occurred at ZT04. Most differentially expressed retinal genes during the light phase were downregulated in the occluded eye. In choroid, the overwhelming number of differentially expressed genes occurred at the end of the light phase (ZT12); most of these were upregulated in the occluded eye (Table 2).
Comparing myopia onset to myopia progression (S8 Table), most differentially expressed genes in retina occurred during the light phase (ZT00 through ZT12) and were downregulated in both conditions. In choroid, many more genes were differentially expressed throughout the day at myopia onset than during progression. Based on findings during myopia onset and/or progression, experimental studies of myopia pathogenesis thus need to include sampling time and myopia status when assaying for gene expression alterations because both can influence findings and bias the outcomes.
Only the current report and our prior report on myopia onset [18] incorporate time of day in sufficient detail to guide future research, particularly on circadian influences. Based on the statistical criteria used to identify differentially expressed genes for both tissues, S9A (retina) and S9B (choroid) Table describe all genes that 1) were differentially expressed at myopia onset; 2) were common to both onset and progression; and 3) were differentially expressed during myopia progression. Comparisons between onset and progression are discussed further below.
OcclVopen*time interaction category
Modeling the interaction of treatment (i.e., occluded vs. open eye gene expression) with time provided the OcclVopen*time interaction categories (Fig 2 and Tables 4 and S4A and S4B). While more differentially expressed genes were identified from inter-eye comparisons at specific times in each tissue (Tables 2 and S3A and S3B), the OcclVopen*time interaction category incorporates data from many times and provides more robust comparisons of differentially regulated genes than the results of separately analyzed time points. We highlight here genes meeting the statistical cut-off of p-adj<0.1 (Table 4), even though broader statistical criteria were needed to generate enough genes to produce the groupings of Fig 2 (see Results text).
Retinal genes in the OcclVopen*time interaction category in myopia progression (Fig 2 and Tables 4 and S4A) include the circadian clock genes PER2 and ARNTL [29] and PCSK1, whose product processes prohormone and neuropeptide precursors [53]. The VIP gene conforms with the long-imputed role of VIP peptide in myopia; but the downregulation of VIP in progressing chick form deprivation myopia here and in chick myopia following the wearing of a minus spectacle lens (i.e., lens-induced myopia) [54] contrasts with its upregulation in primate and guinea pig myopia [55, 56], perhaps representing a species difference. Dermatan sulfate epimerase, a rate limiting enzyme in the synthesis of dermatan sulfate which has a role in tissue development and homeostasis [57, 58] also appears in this category. The protein product of NPNT, an extracellular matrix protein, regulates cell adhesion, differentiation and other functions [59]. ANGPT2 functions to influence vascular modeling and, at least in mammals, may impact development of the aqueous humor outflow pathway [60]; thus, NPNT may point toward a mechanism underlying the link between myopia and glaucoma [6]. Because the retina is a neural tissue, the expression of numerous genes linked to cardiac muscle (MYBPC3, ACTC1, MYL3, MYH7, MYH7B) or to skeletal muscle (MYL1, SMYD1, TNNI2, MYH1B, ACTA1, TNNT3) presents a conundrum for explaining a mechanism for myopia progression.
The choroid contained fewer genes in the OcclVopen*time interaction category during myopia progression than the retina (Fig 2 and Tables 4 and S4B). To highlight a few, the product of PRKCA is a protein kinase that influences proliferation, cell growth, inflammation, neurotransmitter release, circadian rhythmicity and phase setting [61]. While a role for fibroblast growth factor 4 (whose receptor is the product of FGFR4) is not known, other growth factors have been implicated in myopia, including fibroblast growth factor 2 [62]. Much work over centuries has implicated cholinergic signaling and cholinergic receptors in myopia [8, 9, 63], including the M4 muscarinic receptor subtype [64] which is the product of the ChRM4 gene in chick. The protein product of THBS4 influences cell-cell and cell-matrix interactions, tissue remodeling, synaptogenesis, angiogenesis and fibrosis among other activities [65]. FXYD6 encodes phosphohippolin, a modulator of Na,K-ATPase activity [66]; a potential control of eye growth and refraction by ion movements in the retina and choroid has previously been suggested [67, 68].
In the OcclVopen*time interaction category, the retina had several differentially expressed genes common to both myopia onset and progression (S9A and S10 Tables) [18], with both lists restricted to the p-adj<0.1 criterion. MAFF is a transcription factor. R-spondins enhance Wnt signaling [69], and Wnt signaling seemingly impacts myopia in humans and mice [70]. Most of the other common genes comparing OcclVopen*time category in retina have been discussed above. In choroid, only one gene (FXYD6, also discussed above) was shared in the interaction category of both conditions (S9B and S10 Tables). [18] In comparing the heatmaps during myopia progression (S1A and S1B Fig) with the heatmaps of the occlVopen*time interaction category at myopia onset [18], the common differentially expressed genes (S10 Table) between onset and progression tend to fluctuate less dramatically than some of the other the occlVopen*time genes. Whether these characteristics have mechanistic significance is uncertain.
Overall category
The overall category lists the genes where the occluded vs. open eye comparisons were similar statistically in magnitude and direction throughout the day even though a statistically significant difference was not required at each sampling time. As a result of the statistical modeling, some genes were identified in each tissue for both overall and OcclVopen*time interaction categories. Despite the apparent statistical ambiguities in the overlap between these two categories, identification of these genes nonetheless has potential value in studying myopia pathogenesis. Only 14 genes in retina and 2 genes in choroid fell into the overall category during myopia progression. While only a small number of genes, their protein products may influence refractive development throughout the day.
Of the genes in the overall category, those in retina seemingly are most informative (Table 3 and S3A, S8 and S9A). To highlight a few besides transcription factors, SPR participates in the biosynthesis of tetrahydrobiopterin, a cofactor for tyrosine hydroxylase and tryptophan hydroxylase that have roles in dopamine and serotonin synthesis as well as functioning as an antioxidant [71]; dopamine and serotonin have each been implicated in myopia pathogenesis [72–74]. The protein products of MIDN and LONRF1 participate in ubiquination and protein degradation, among other potential actions [75, 76]. Two genes encode multifunctional proteins that interact negatively with mitogen-activated protein (MAP) kinase signaling: SPRY4, a protein inhibitor of MAP kinase signaling [77]; and DUSP4, a phosphatase acting at MAP kinase as well as at other proteins [78, 79]. MAP kinases are components of signaling pathways that govern general developmental phenomena such as cell proliferation and cell differentiation [80]. In the choroid, cartilage oligomeric matrix protein is the only gene product identified of the two differentially regulated genes in the overall category (S3B and S9B Tables); this large, complex extracellular protein exerts a variety of functions in the extracellular matrix [81].
The overall categories in the differentially expressed genes during myopia onset [18] and progression share only three genes in retina: LONRF1, RAD54L2 and DUSP4 (S9A Table). In choroid, no differentially regulated genes were common to both myopia onset and progression (S9B Table).
Other differentially expressed genes at specific times
While less robust statistically than the two categories just discussed, additional genes identified at specific times merit some comment. Each tissue developed either up-regulated or down-regulated genes at many but not all times of day during myopia progression (Tables 2 and S3A and S3B). The overwhelming majority of differentially expressed genes in choroid occurred at ZT12 and most differentially expressed genes in retina occurred at earlier times, consistent with the notion of a time delay between at least some retinal signals and their actions at the choroid.
As examples for retina at ZT04 and ZT16 (Tables 2 and S3A), NOG encodes an inhibitor of bone morphogenetic proteins (BMPs); both NOG and BMPs have been implicated in myopia [22, 35]. GJD2, also implicated in human myopia [82], produces a gap junction protein connexin-36 that is regulated by dopamine [83, 84]. DIO2 has been previously associated with myopia [22], as have genes encoding urotensin (e.g., UTS2B) [22, 54]. The protein product of ANGP2 (angiopoitin-2) impacts vascular growth and inflammation [85], and angiopoitin-2 is a new target for retinal vascular disease [86]. NTNT encodes nephronectin, an extracellular matrix protein involved in cell adhesion and cell differentiation [59]. How, or even if, these protein products impact myopia requires further study.
In the choroid, the many genes differentially expressed at ZT12 (Tables 2 and S3B) during myopia progression impact such processes as inflammation, membrane signaling, ion channels and inflammation. While genes involving neural signaling and vision might have been expected to be found in retina, they here are found instead in choroid. Included in these genes are numerous genes related to receptors for the neurotransmitters GABA, glutamate and acetylcholine, several opsin genes and PRPH2 (also known as retinal degeneration slow, or RDS). Efforts during tissue dissection sought to minimize cross-contamination to the extent practical between the retina and choroid (see Methods). The genes for IGF-2 and several FGF receptor subtypes are identified. Also identified are DJD2 (seen in retina, above) and DJC2, both of which encode gap junction proteins [82]. The circadian clock gene PER3 and ASMT (whose protein product catalyzes the final step in melatonin synthesis) relate directly to potential circadian dysregulation in myopia, the hypothesis underlying this study.
Considering only the numbers of differentially regulated genes (S8 Table), there are surprisingly few common genes comparing myopia onset to myopia progression. Perhaps, there are relatively few signaling mechanisms common to both myopia onset and progression, and the common genes could be identifying common mechanisms essential to the characteristic ocular elongation in myopia (S9A and S9B Table). The limited overlap in the overall gene categories suggest that these genes and the processes they regulate (e.g., protein degradation or MAP kinase signaling in choroid) may be active throughout the day, and that only a few genes underlie time-independent signaling common to myopia onset and progression. Perhaps also, the distinct differentially expressed genes at myopia onset or during myopia progression reveal useful directions to favorably impact myopia, suggesting different interventions specifically for reducing the onset or the progression of myopia. Substantiating any of these notions, however, will require further research.
Occluded vs. open eye differences at multiple times
Only a few individual genes developed statistically significant occluded vs. open eye differences at more than one time during myopia progression (Fig 3 and Tables 5 and S5). Individual genes may possibly have more influence on refraction if differentially expressed at multiple times in the day. Because of the 4-hour interval between tissue sampling times, however, gene expression differences at consecutive times may develop from a more pronounced effect at an intermediate time, rather than indicating a prolonged action. In retina, only 7 genes were differentially expressed at multiple times: of these, 3 were identified at non-consecutive times. In choroid, only 3 genes showed expression differences at multiple times: of these, 1 was identified at non-consecutive times (Fig 3 and Tables 5 and S5). The limited number of genes differentially expressed in each tissue at multiple times likely follows the concentration of differentially expressed genes to specific times (Table 2). Perhaps, the most interesting genes for the pathogenesis of myopia progression are PCSK1, DUSP and VIP in retina; each is discussed above. Many more genes were differentially expressed at multiple times in our previous study on myopia onset: 30 in retina and 89 in choroid [18]. Comparing the present progression study to myopia onset [18], only 5 genes were common to both studies in retina; and only one, in choroid (S11 Table). Like the broad perspective above on the limited common genes at myopia onset and during myopia progression, the limited common genes found at multiple times during myopia progression supports the hypothesis that myopia onset and progression may have different mechanisms.
GSEA pathways
Of the many signaling, structural and metabolic pathways identified, those from the occlVopen*time and overall categories will be emphasized below (S7A and S7B Table). During myopia progression, pathways in retina in the occlVopen*time category enriched in occluded or open eyes include several pathways related to muscle, extracellular matrix, autophagy and the ARENRF2 anti-oxidant/detoxifying pathway [87]. The overall category in retina includes a number of pathways addressing oxidative metabolism, transport, protein processing, and nerve growth factor. Like individual genes, pathways in the choroid during myopia progression enriched in occluded or open eyes include many visual system and neural signaling pathways. In the occlVopen*time category, these include the cone, rhodopsin and phototransduction pathway, numerous neurotransmitter-related pathways (e.g., GABA, glutamate, norepinephrine, serotonin and synaptic vesicles), plus some involved in inflammation (e.g., interleukin 12) or cell biology. The overall category includes pathways enriched in occluded or open eyes for cones, phototransduction, some growth factors (IGF1, NGF), cell function, structure and inflammation.
Pathways identified previously at myopia onset are both similar and different from those found during myopia progression, using the same statistical criteria [18]. Retinal pathways found at myopia onset for the occlVopen*time category contrast with those during progression and include the rhodopsin pathway, Krebs cycle pathway and the mitochondrial complex 1 assembly pathway. In the overall category, retinal pathways at myopia onset include those involved in rhodopsin function, phototransduction, oxidative metabolism and various structural components. More similar to progressing myopia, the choroidal pathways at myopia onset for the occlVopen*time category contain neurotransmitter-related pathways (e.g., serotonin, dopamine, glutamate, norepinephrine, GABA, acetylcholine, etc.), cone pathways, metabolic pathways, as well as a few others. Among the choroidal pathways in the overall category at myopia onset are those related to neurotransmitters (e.g., GABA, serotonin, acetylcholine, GABA), cones, rhodopsin, visual transduction, ion channels, inflammation, etc. The pathway differences, especially for retina, further support the notion of substantial differences in the mechanisms of myopia onset and progression.
Relative to the circadian dysregulation hypothesis underlying this investigation, the circadian clock pathway is detected during myopia progression in retina at ZT04 and ZT20 but not in choroid; the clock pathway is not identified in either the occlVopen*time or overall categories of either tissue. The pathways from the occlVopen*time interaction and overall categories associated with photoreceptors and photoreceptor biology correspond to findings implicating photoreceptors in refractive error development in laboratory animals [67, 88] and in human genetics [16, 19, 89]. As in myopia onset [18], the many neurotransmitter-related pathways found in choroid during myopia progression suggest complex neuronal mechanisms in this tissue contributing not only to myopia onset but also to myopia progression. Besides our results in myopia onset and myopia progression, prior genome-wide assays of the retina have identified many complex pathways in chick [20–22, 90, 91] or mammalian [23, 24, 92] myopia. While a necessary consequence of the statistical nature of pathway identifications, the large number and variability of GSEA-generated pathways emphasize the complexity of the mechanisms underlying myopia onset and progression of established myopia.
Human vs. chick genes.
In either retina or choroid, we found only a few differentially expressed genes represented during myopia progression among the many candidate genes for human myopia (Table 6). This limited overlap parallels our prior results at myopia onset [18] and the limited overlap in another study comparing chick retinal genes to those in human myopia at 4 hours and 24 hours of form deprivation [20]. The reason(s) for this limited overlap between chick and human myopia genes is uncertain. The anatomical parallels between form-deprivation myopia and human myopia, the induction of myopia in children from conditions that interfere with the visual axis (e.g., eyelid ptosis or corneal scarring) and similar anti-myopia responses to muscarinic antagonists in chicks and children [9, 93] suggest at least some similarity in underlying mechanisms. While requiring direct study, those human candidate genes overlapping with chick myopia genes may somehow be essential to myopia pathogenesis independent of species.
Among the retinal genes during chick myopia progression overlapping with human myopia genes in retina, EHPA2 encodes one of the Eph receptors, a large family of receptor tyrosine kinases that mediate many processes, such as cell-to-cell communication and cell attachment [94], as well as several phosphodiesterases and DJD2, a gap junction gene [82]. In the choroid, overlapping genes between chick and human include ECEL1, DJD2, PAX6, BMP3, FGFR3 and GDF11, all genes potentially interacting with signaling or growth mechanisms. The protein products of several genes have roles in neuronal biology signaling: CTNND2, GABRR1 and GRIN2B. Surprisingly, genes associated with retinal disorders (CLUL1, VSX1, RD3, RS1, and CABP4) and a cone pigment (OPN1LW) were identified in choroid but not retina of progressing chick myopia (Table 6). From single cell transcriptomics, several opsin genes also were recently found to be expressed in one or more choroidal cell types [95]. Refractive errors are associated with retinal diseases [89, 96], but the choroidal expression of these genes will require more study to understand. Refractive errors also occur in ocular diseases affecting other components of the eye, such as cornea or lens [1, 97], but whether our findings in retina or choroid relate to these conditions requires direct investigation.
As in our study of myopia onset [18], the overlap of human candidate myopia genes with chick myopia genes at discrete times in retina and in choroid during myopia progression suggests some type of connection to circadian biology that motivated this study. In both retina and choroid, some overlapping chick-human genes occur only during myopia onset, only during myopia progression or during both conditions. Perhaps, these distinctions between genes may help to identify mechanisms active at myopia onset, active at myopia progression or general pro-myopia mechanisms. As examples, the BMP’s and GABA signaling may be part of general myopia mechanisms, but the gap junction protein product of GJD2 may impact myopia progression. Such distinctions accessible in laboratory myopia models are hard to discern in clinical genetic risk factor studies. Such hypotheses would benefit from an improved understanding of the relationship between form-deprivation myopia and common human myopia.
Circadian biology and clinical myopia
Many genes and pathways are linked by genetic studies to human myopia, including those involved with circadian rhythms [16]. Sleep disturbances, a potential indicator of circadian disruption, can occur in myopic children [98–100] but are not consistently detected [101–105]; the type of the sleep disorder also varies between studies [100, 106–108]. Sleep quality also may impact orthokeratology, a treatment for myopia [109]. The clock gene PER3 was identified as a rare variant in human myopia [49] (Table 6) and was found in the choroid during myopia progression at ZT12 (S3B Table); PER3 is involved in sleep regulation, non-visual responses to light and a variety of circadian responses [110], supporting potential roles for light and circadian biology in myopia. Increased time outdoors associates with a lower myopia prevalence but does not clearly impact myopia progression [111], suggesting perhaps that light and/or circadian effects may differently impact myopia onset and progression. Seeking a direct circadian effect on myopia, studies of children, early teenagers and young adults have so far resulted in conflicting findings of an association of myopia with systemic melatonin, an intrinsic mediator of circadian rhythms [112–114]. Reinforced by contemporary human genetics [16], environmental lighting has long been postulated to impact refractive development [15, 115–118]. The worsening problem of light pollution from ambient artificial light at night [119] can disrupt circadian rhythms and has been linked to sleep disorders in children [120] and perhaps myopia [121]. With recent concerns about putative pro-myopia effects of light from electronic screens [122], further investigation is needed to unravel the roles of light and sleep in myopia pathogenesis [123].
Besides the few common genes, many specific genes have previously been found and postulated as potential mediators of myopia both in experimental animals and in children [16, 19]. These many potential mediators of myopia onset and/or progression greatly complicate designing laboratory and clinical studies to study myopia mechanisms and to develop novel therapies. Perhaps because of the increasing biochemical and pharmacological complexity, novel myopia therapies increasingly assess specific wavelengths of light exposure, outdoor activities, optical devices and behavioral modification. Despite the clinical interest, only modest anti-myopia efficacy has resulted so far [124] with variable rebound of myopia on terminating these treatments [125].
Extending the results on the onset of form-deprivation myopia in chick, the current study finds a key role for time of day in potential retinal and choroidal signaling mechanisms during the progression of experimental myopia and further buttresses the notion that circadian biology exerts a central role in the mechanism(s) of myopia. Most available studies of myopia pathogenesis in children inadequately assess any circadian disorder or time of day, including the numerous light, optical or behavioral remedies proposed as potential myopia therapies. Identifying the type of any circadian disorder(s) causing/augmenting myopia and establishing any role for time of day in myopia pathogenesis would seem essential in clinical studies seeking the cause of myopia and improved anti-myopia interventions.
Supporting information
S1 Fig. A Retinal heatmap, myopia progression, occlVopen*time interaction.
Heatmap of the log2 fold-changes in retina are shown across the sampling times for the occluded vs. open eye differences interacting with time (occlVopen*time interaction) during myopia progression. Key in upper right gives the magnitude of log2 fold-changes. The color indicates the direction of occluded vs. open eye gene expression differences–red, expression higher in occluded eye; blue, expression lower in the occluded eye. See Fig 2 and Tables 4 and S4A. lfc; log2 fold-change, preceded by the ZT of the tissue sampling time. Left ordinate: dendrogram relating the gene clusters. Right ordinate: gene name. B. Choroidal heatmap, myopia progression, occlVopen*time interaction. Heatmap of the log2 fold-changes in retina are shown across the sampling times for the occluded vs. open eye differences interacting with time (occlVopen*time interaction) during myopia progression. Key in upper right gives the magnitude of log2 fold-changes. The color indicates the direction of occluded vs. open eye gene expression differences–red, expression higher in occluded eye; blue, expression lower in the occluded eye. See Fig 2, Tables 4 and S4B. lfc; log2 fold-change, preceded by the ZT of the tissue sampling time. Left ordinate: dendrogram relating the gene clusters. Right ordinate: gene name.
https://doi.org/10.1371/journal.pone.0307091.s001
(ZIP)
S1 Table. RNA quantity and quality for each sample.
The total amount of RNA extracted from each sample and two quality measures are shown for each chick, eye, tissue and ZT sampling time. chick ID, an arbitrary identification number used for this experimental series. ZT, Zeitgeber time of tissue sampling, in hours. RIN, RNA integrity number. Mean Quality Score, or Phred quality score (Q score) to assess accuracy of a sequencing platform.
https://doi.org/10.1371/journal.pone.0307091.s002
(XLSX)
S2 Table. Variability of gene expression across time during myopia progression.
The variability of expression levels across time for all genes with measurable p-adj values during myopia progression, ordered by p-adj values, are shown for S2A Table) retinas of occluded eyes; S2B Table) retinas of contralateral open eyes; S2C Table) choroids of occluded eyes; and S2D Table) choroids of contralateral open eyes. baseMean, average normalized counts over all samples. p-adj, p-value corrected for the false discovery rate (FDR) with the Benjamini-Hochberg method.
https://doi.org/10.1371/journal.pone.0307091.s003
(XLSX)
S3 Table.
A. Comparisons of occluded to open eyes over time during myopia progression, retina. The gene expression levels in occluded vs. contralateral open eyes over time in retina are provided for genes with between-eye differences meeting the criterion of p-adj<0.05, sorted by log2 fold-change. ZT, Zeitgeber time of tissue sampling, in hours. overall, retinal genes that differed statistically between occluded vs. contralateral open eyes over the entire day, identified by assessing all replicates at all time points simultaneously (see text, Table 2). lfcSE, standard error of the log2 fold-change. p-adj, p-value corrected for the false discovery rate (FDR) with the Benjamini-Hochberg method. B. Comparisons of occluded to open eyes over time during myopia progression, choroid. The gene expression levels in occluded vs. contralateral open eyes over time in choroid are provided for genes with between-eye differences meeting the criterion of p-adj<0.05, sorted by log2 fold-change. ZT, Zeitgeber time of tissue sampling, in hours. overall, retinal genes that differed statistically between occluded vs. contralateral open eyes over the entire day, identified by assessing all replicates at all time points simultaneously (see text, Table 2). lfcSE, standard error of the log2 fold-change. p-adj, p-value corrected for the false discovery rate (FDR) with the Benjamini-Hochberg method.
https://doi.org/10.1371/journal.pone.0307091.s004
(ZIP)
S4 Table.
A. Group assignments of occlVopen*time interactions in retina during myopia progression. The inter-eye gene expression differences that interacted with time (i.e., occlVopen*time interaction), meeting the broadened statistical criteria described in the Results text, along with their cluster (group) assignments are shown for retina. The genes are ranked by p-adj values. p-adj, p-value corrected for the false discovery rate (FDR) using the Benjamini-Hochberg method. cluster, see Fig 2 for the representation of the patterns of the clusters (i.e., groups) and the text for the method of determining the occlVopen*time interactions. B. Group assignments of occlVopen*time interactions in choroid during myopia progression. The inter-eye gene expression differences that interacted with time (i.e., occlVopen*time interaction), meeting the broadened statistical criteria described in the Results text, along with their cluster (group) assignments are shown for choroid. The genes are ranked by p-adj values. p-adj, p-value corrected for the false discovery rate (FDR) using the Benjamini-Hochberg method. cluster, see Fig 2 for the representation of the patterns of the clusters (i.e., groups) and the text for the method of determining the occlVopen*time interactions.
https://doi.org/10.1371/journal.pone.0307091.s005
(ZIP)
S5 Table. Gene expression differences between occluded and open eyes at more than one time during myopia progression.
The Venn Diagrams of Fig 3 provide the genes with between-eye expression differences with p-adj<0.05 at more than one time. The S5 Table lists the ZT times, the number and names of genes differentially expressed a more than one ZT time, the gene descriptions, the directions of gene expression changes and the log2 fold changes at statistically significant times. Retina, above; choroid, below. ZT, Zeitgeber time of tissue sampling, in hours.
https://doi.org/10.1371/journal.pone.0307091.s006
(DOCX)
S6 Table.
A. Complete data for occluded vs. open eye comparisons in retina during myopia progression. The tables provide occluded vs. contralateral open eye gene expression differences over time in retina for all genes with the log2 fold-changes of the differences, sorted by p-adj values. ZT, Zeitgeber time of tissue sampling, in hours. overall, genes that differed statistically between occluded vs. contralateral open eyes over the entire day in retina, identified by assessing all replicates at all time points simultaneously (see text).baseMean, average normalized counts across all samples. lfcSE, standard error of the log2 fold-change. p-adj, p-value corrected for the false discovery rate (FDR) using the Benjamini-Hochberg method. B. Complete data for occluded vs. open eye comparisons in choroid during myopia progression. The tables provide occluded vs. contralateral open eye gene expression differences over time in choroid for all genes with the log2 fold-changes of the differences, sorted by p-adj values. ZT, Zeitgeber time of tissue sampling, in hours. overall, genes that differed statistically between occluded vs. contralateral open eyes over the entire day in choroid, identified by assessing all replicates at all time points simultaneously (see text). baseMean, average normalized counts across all samples. lfcSE, standard error of the log2 fold-change. p-adj, p-value corrected for the false discovery rate (FDR) using the Benjamini-Hochberg method.
https://doi.org/10.1371/journal.pone.0307091.s007
(ZIP)
S7 Table.
A. GSEA pathways in retina during myopia progression. Pathway assignments were based on the relative expression of retinal genes in occluded eyes vs. contralateral open eyes. Two general types of GSEA pathways were identified: 1) pathways enriched in occluded eyes relative to the contralateral open genes; and 2) pathways enriched in open eyes relative to the contralateral occluded eyes, corresponding to decreased expression in occluded eyes. The genes generating the pathways included those identified for specific times of the day, the “overall” category (see S3A and S3B Table), and the occlVopen*time interaction (see Tables 4 and S4A and S4B). Pathways listed met the statistical criterion of FDR q-value<0.05. GSEA, Gene Set Enrichment Analysis. ZT, Zeitgeber time of tissue sampling, in hours. NES, normalized enrichment score. FDR, false discovery rate. B. GSEA pathways in choroid during myopia progression. Pathway assignments were based on the relative expression of choroidal genes in occluded eyes vs. contralateral open eyes. Two general types of GSEA pathways were identified: 1) pathways enriched in occluded eyes relative to the contralateral open genes; and 2) pathways enriched in open eyes relative to the contralateral occluded eyes, corresponding to decreased expression in occluded eyes. The genes generating the pathways included those identified for specific times of the day, the “overall” category (see S3A and S3B Table), and the occlVopen*time interaction (see Tables 4 and S4A and S4B). Pathways listed met the statistical criterion of FDR q-value<0.05. GSEA, Gene Set Enrichment Analysis. ZT, Zeitgeber time of tissue sampling, in hours. NES, normalized enrichment score. FDR, false discovery rate.
https://doi.org/10.1371/journal.pone.0307091.s008
(ZIP)
S8 Table. Myopia onset vs. myopia progression: Gene numbers with expression differences between occluded and open eyes.
Myopia progression vs. myopia onset, summary of gene numbers with expression differences between occluded and open eyes at each time (p-adj<0.05). “UP” = occluded/open eyes: + fold-change; “DOWN” = occluded/open eyes: − fold-change. ZT, Zeitgeber time of tissue sampling, in hours. “overall” = genes with occluded vs. open eye differences that were similar in magnitude and direction at all time points (see text). S3A (retina) and S3B (choroid) Table of the current study and S2A (retina) and S2B (choroid) Table of the myopia onset study [18] include by time the gene names and gene descriptions for each tissue with expression differences (p-adj<0.05), ranked by log2 fold change.
https://doi.org/10.1371/journal.pone.0307091.s009
(DOCX)
S9 Table.
A. Distinct and common genes in retina, myopia onset vs. myopia progression. Distinct and common genes in retina for the occlVopen*time interaction category, for the overall category and for each sampling time for occluded vs. open gene expression comparisons for myopia onset and myopia progression. Data derive from S2 and S3 Tables in our study of myopia onset [18] and from S3A, S3B, S4A and S4B Tables of current study of myopia progression. The statistical criteria for selecting genes were identical for the two studies: p-adj<0.1 for the occlVopen*time interaction categories, and p-adj<0.05 for the overall category and for the individual times. ZT, Zeitgeber time of tissue sampling, in hours. overall, genes that differed statistically between occluded vs. contralateral open eyes over the entire day. baseMean, average normalized counts across all samples. lfcSE, standard error of the log2 fold-change. p-adj, p-value corrected for the false discovery rate (FDR) using the Benjamini-Hochberg method. B. Distinct and common genes in choroid, myopia onset vs. myopia progression. Distinct and common genes in choroid for the occlVopen*time interaction category, for the overall category and for each sampling time for occluded vs. open gene expression comparisons for myopia onset and myopia progression. Data derive from S2 and S3 Tables in our study of myopia onset [18] and from S3A, S3B, S4A and S4B Tables of current study of myopia progression. The statistical criteria for selecting genes were identical for the two studies: p-adj<0.1 for the occlVopen*time interaction categories, and p-adj<0.05 for the overall category and for the individual times. ZT, Zeitgeber time of tissue sampling, in hours. overall, genes that differed statistically between occluded vs. contralateral open eyes over the entire day. baseMean, average normalized counts across all samples. lfcSE, standard error of the log2 fold-change. p-adj, p-value corrected for the false discovery rate (FDR) using the Benjamini-Hochberg method.
https://doi.org/10.1371/journal.pone.0307091.s010
(ZIP)
S10 Table. Myopia onset vs. myopia progression: Genes with OcclVopen*time interactions over the full day.
Genes with occluded vs. open eye differences interacting with time over the full day (i.e., the occlVopen*interaction category) from Table 3 here and from Table 3 of our study of myopia onset [18]). Genes met the criterion of p-adj<0.1 in both studies (see texts). Named genes are listed, and Ensemble gene ID numbers are provided for non-named genes. See S9A (retina) and S9B (choroid) Table for more information including gene descriptions. Retina, top; choroid, bottom. Blue shading = genes that are common to myopia onset and myopia progression. * = the only non-named gene common to progression and onset, which appears in retina, is described as “SHC adaptor protein 4”.
https://doi.org/10.1371/journal.pone.0307091.s011
(DOCX)
S11 Table. Myopia onset vs. myopia progression: Genes with occluded vs. open eye differences at multiple times.
Common genes with occluded vs. open eye differences at more than one time from the current report on myopia progression and from our study of myopia onset [18] for genes meeting the criterion of p-adj<0.05 in each study. For each gene, only times meeting this statistical criterion are shown. Retina, top; choroid, bottom. ZT, Zeitgeber time of tissue sampling, in hours.
https://doi.org/10.1371/journal.pone.0307091.s012
(DOCX)
Acknowledgments
We thank Drs. Tapan Ganguly and Jonathan Schug of the Penn Genomics and Sequencing Core for assistance in preparing RNA from the tissues and Azenta Life Sciences (South Plainfield, NJ, USA) for processing and sequencing the RNA.
References
- 1. Curtin BJ. The Myopias. Philadelphia: Harper & Row; 1985.
- 2. Koh V, Yang A, Saw SM, Chan YH, Lin ST, Tan MM, et al. Differences in prevalence of refractive errors in young Asian males in Singapore between 1996–1997 and 2009–2010. Ophthalmic Epidemiol. 2014;21(4):247–55. pmid:24990474
- 3. Vitale S, Sperduto RD, Ferris FL, 3rd. Increased prevalence of myopia in the United States between 1971–1972 and 1999–2004. Arch Ophthalmol. 2009;127(12):1632–9. pmid:20008719
- 4. Williams KM, Verhoeven VJ, Cumberland P, Bertelsen G, Wolfram C, Buitendijk GH, et al. Prevalence of refractive error in Europe: the European Eye Epidemiology (E(3)) Consortium. Eur J Epidemiol. 2015;30(4):305–15. pmid:25784363
- 5. Holden BA, Fricke TR, Wilson DA, Jong M, Naidoo KS, Sankaridurg P, et al. Global Prevalence of Myopia and High Myopia and Temporal Trends from 2000 through 2050. Ophthalmology. 2016;123(5):1036–42. pmid:26875007
- 6. Sankaridurg P, Tahhan N, Kandel H, Naduvilath T, Zou H, Frick KD, et al. IMI Impact of Myopia. Invest Ophthalmol Vis Sci. 2021;62(5):2. pmid:33909036
- 7. Huang Y, Chen X, Zhuang J, Yu K. The Role of Retinal Dysfunction in Myopia Development. Cell Mol Neurobiol. 2022. pmid:36427109
- 8.
Stone RA. Chap 9. Myopia pharmacology: etiologic clues, therapeutic potential. In: Yorio T, Clark A, Wax M, editors. Ocular therapeutics: an eye on new discoveries. New York: Elsevier/Academic Press; 2008. p. 167–96.
- 9. Stone RA, Pardue MT, Iuvone PM, Khurana TK. Pharmacology of myopia and potential role for intrinsic retinal circadian rhythms. Exp Eye Res 2013;114:35–47. pmid:23313151
- 10. Wallman J. Retinal control of eye growth and refraction. Progress in Retinal Research. 1993;12:133–53.
- 11. Smith EL, 3rd, Hung LF, Arumugam B. Visual regulation of refractive development: insights from animal studies. Eye (Lond). 2014;28(2):180–8.
- 12. Liu Y, Wang L, Xu Y, Pang Z, Mu G. The influence of the choroid on the onset and development of myopia: from perspectives of choroidal thickness and blood flow. Acta Ophthalmol. 2021;99(7):730–8. pmid:33550704
- 13. Nickla DL. Ocular diurnal rhythms and eye growth regulation: where we are 50 years after Lauber. Exp Eye Res. 2013;114:25–34. pmid:23298452
- 14. Nickla DL, Wallman J. The multifunctional choroid. Prog Retin Eye Res. 2010;29(2):144–68. pmid:20044062
- 15. Chakraborty R, Ostrin LA, Nickla DL, Iuvone PM, Pardue MT, Stone RA. Circadian rhythms, refractive development, and myopia. Ophthalmic Physiol Opt. 2018;38(3):217–45. pmid:29691928
- 16. Hysi PG, Choquet H, Khawaja AP, Wojciechowski R, Tedja MS, Yin J, et al. Meta-analysis of 542,934 subjects of European ancestry identifies new genes and mechanisms predisposing to refractive error and myopia. Nat Genet. 2020;52(4):401–7. pmid:32231278
- 17. Li L, Yu Y, Zhuang Z, Wu Q, Lin S, Hu J. Circadian rhythm, ipRGCs, and dopamine signalling in myopia. Graefes Arch Clin Exp Ophthalmol. 2023. pmid:37864638
- 18. Stone RA, Tobias JW, Wei W, Schug J, Wang X, Zhang L, et al. Diurnal retinal and choroidal gene expression patterns support a role for circadian biology in myopia pathogenesis. Sci Rep. 2024;14(1):533. pmid:38177383
- 19. Tedja MS, Wojciechowski R, Hysi PG, Eriksson N, Furlotte NA, Verhoeven VJM, et al. Genome-wide association meta-analysis highlights light-induced signaling as a driver for refractive error. Nat Genet. 2018;50(6):834–48. pmid:29808027
- 20. Karouta C, Kucharski R, Hardy K, Thomson K, Maleszka R, Morgan I, et al. Transcriptome-based insights into gene networks controlling myopia prevention. Faseb j. 2021;35(9):e21846. pmid:34405458
- 21. Riddell N, Giummarra L, Hall NE, Crewther SG. Bidirectional Expression of Metabolic, Structural, and Immune Pathways in Early Myopia and Hyperopia. Front Neurosci. 2016;10:390. pmid:27625591
- 22. Stone RA, McGlinn AM, Baldwin DA, Tobias JW, Iuvone PM, Khurana TS. Image defocus and altered retinal gene expression in chick: clues to the pathogenesis of ametropia. Invest Ophthalmol Vis Sci. 2011;52(8):5765–77. pmid:21642623
- 23. Tkatchenko TV, Shah RL, Nagasaki T, Tkatchenko AV. Analysis of genetic networks regulating refractive eye development in collaborative cross progenitor strain mice reveals new genes and pathways underlying human myopia. BMC Med Genomics. 2019;12(1):113. pmid:31362747
- 24. Tkatchenko TV, Tkatchenko AV. Genome-wide analysis of retinal transcriptome reveals common genetic network underlying perception of contrast and optical defocus detection. BMC Med Genomics. 2021;14(1):153. pmid:34107987
- 25. Stone RA, McGlinn AM, Chakraborty R, Lee DC, Yang V, Elmasri A, et al. Altered ocular parameters from circadian clock gene disruptions. PLoS One. 2019;14:e0217111. pmid:31211778
- 26. Chakraborty R, Landis EG, Mazade R, Yang V, Strickland R, Hattar S, et al. Melanopsin modulates refractive development and myopia. Exp Eye Res. 2022;214:108866. pmid:34838844
- 27. Liu AL, Liu YF, Wang G, Shao YQ, Yu CX, Yang Z, et al. The role of ipRGCs in ocular growth and myopia development. Sci Adv. 2022;8(23):eabm9027. pmid:35675393
- 28. Chakraborty R, Collins MJ, Kricancic H, Davis B, Alonso-Caneiro D, Yi F, et al. The effect of intrinsically photosensitive retinal ganglion cell (ipRGC) stimulation on axial length changes to imposed optical defocus in young adults. J Optom. 2023;16(1):53–63. pmid:35589503
- 29. Stone RA, Wei W, S S, McGeehan B, Engelhart KC, Khurana TS, et al. Visual image quality impacts circadian rhythm-related gene expression in retina and choroid − a potential mechanism for ametropias. Invest Ophthalmol Vis Sci 2020;61:doi:https://doi.org/https://doi.org/10.1167/iovs.1.5.13.
- 30. Mure LS, Le HD, Benegiamo G, Chang MW, Rios L, Jillani N, et al. Diurnal transcriptome atlas of a primate across major neural and peripheral tissues. Science. 2018;359(6381). pmid:29439024
- 31. Zhang R, Lahens NF, Ballance HI, Hughes ME, Hogenesch JB. A circadian gene expression atlas in mammals: implications for biology and medicine. Proc Natl Acad Sci U S A. 2014;111(45):16219–24. pmid:25349387
- 32. Troilo D, Smith EL, 3rd, Nickla DL, Ashby R, Tkatchenko AV, Ostrin LA, et al. IMI ‐ Report on Experimental Models of Emmetropization and Myopia. Invest Ophthalmol Vis Sci. 2019;60(3):M31–m88. pmid:30817827
- 33. Wallman J, Turkel J, Trachtman J. Extreme myopia produced by modest change in early visual experience. Science. 1978;201:1249–51. pmid:694514
- 34. Brown DM, Mazade R, Clarkson-Townsend D, Hogan K, Datta Roy PM, Pardue MT. Candidate pathways for retina to scleral signaling in refractive eye growth. Exp Eye Res. 2022;219:109071. pmid:35447101
- 35. McGlinn AM, Baldwin DA, Tobias JW, Budak MT, Khurana TS, Stone RA. Form deprivation myopia in chick induces limited changes in retinal gene expression. Invest Ophthalmol Vis Sci. 2007;48:3430–6. pmid:17652709
- 36. Morgan IG, French AN, Ashby RS, Guo X, Ding X, He M, et al. The epidemics of myopia: Aetiology and prevention. Prog Retin Eye Res. 2018;62:134–49. pmid:28951126
- 37. Kee CS D. M, Wallman J. Differences in time course and visual requirements of ocular responses to lenses and diffusers. Invest Ophthalmol Vis Sci. 2001;42:575–83. pmid:11222513
- 38. Nickla DL. The phase relationships between the diurnal rhythms in axial length and choroidal thickness and the association with ocular growth rate in chicks. J Comp Physiol A Neuroethol Sens Neural Behav Physiol. 2006;192(4):399–407. pmid:16380843
- 39. Patro R, Duggal G, Love MI, Irizarry RA, Kingsford C. Salmon provides fast and bias-aware quantification of transcript expression. Nat Methods. 2017;14(4):417–9. pmid:28263959
- 40.
R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; 2023.
- 41. Love MI, Soneson C, Hickey PF, Johnson LK, Pierce NT, Shepherd L, et al. Tximeta: Reference sequence checksums for provenance identification in RNA-seq. PLoS Comput Biol. 2020;16(2):e1007664. pmid:32097405
- 42. Durinck S, Spellman PT, Birney E, Huber W. Mapping identifiers for the integration of genomic datasets with the R/Bioconductor package biomaRt. Nat Protoc. 2009;4(8):1184–91. pmid:19617889
- 43. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550. pmid:25516281
- 44. Heberle H, Meirelles GV, da Silva FR, Telles GP, Minghim R. InteractiVenn: a web-based tool for the analysis of sets through Venn diagrams. BMC Bioinformatics. 2015;16(1):169. pmid:25994840
- 45. Pantano I. DEGreport: Report of DEG analysis. R package version 1.34.0 2022 [http://lpantano.github.io/DEGreport/.].
- 46. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102(43):15545–50. pmid:16199517
- 47. Haarman AEG, Klaver CCW, Tedja MS, Roosing S, Astuti G, Gilissen C, et al. Identification of Rare Variants Involved in High Myopia Unraveled by Whole Genome Sequencing. Ophthalmol Sci. 2023;3(4):100303. pmid:37250922
- 48. Liu Y, Zhang JJ, Piao SY, Shen RJ, Ma Y, Xue ZQ, et al. Whole-Exome Sequencing in a Cohort of High Myopia Patients in Northwest China. Front Cell Dev Biol. 2021;9:645501. pmid:34222226
- 49. Musolf AM, Haarman AEG, Luben RN, Ong JS, Patasova K, Trapero RH, et al. Rare variant analyses across multiethnic cohorts identify novel genes for refractive error. Commun Biol. 2023;6(1):6. pmid:36596879
- 50. Su J, Yuan J, Xu L, Xing S, Sun M, Yao Y, et al. Sequencing of 19,219 exomes identifies a low-frequency variant in FKBP5 promoter predisposing to high myopia in a Han Chinese population. Cell Rep. 2023;42(5):112510. pmid:37171956
- 51. Yu X, Yuan J, Chen ZJ, Li K, Yao Y, Xing S, et al. Whole-Exome Sequencing Among School-Aged Children With High Myopia. JAMA Netw Open. 2023;6(12):e2345821. pmid:38039006
- 52. Hughes ME, Abruzzi KC, Allada R, Anafi R, Arpat AB, Asher G, et al. Guidelines for Genome-Scale Analysis of Biological Rhythms. J Biol Rhythms. 2017;32(5):380–93. pmid:29098954
- 53. Chrétien M, Mbikay M. 60 YEARS OF POMC: From the prohormone theory to pro-opiomelanocortin and to proprotein convertases (PCSK1 to PCSK9). J Mol Endocrinol. 2016;56(4):T49–62. pmid:26762158
- 54. Shan SW, Wang PF, Cheung JKW, Yu F, Zheng H, Luo S, et al. Transcriptional profiling of the chick retina identifies down-regulation of VIP and UTS2B genes during early lens-induced myopia. Mol Omics. 2022;18(5):449–59. pmid:35420081
- 55. Stone RA, Laties AM, Raviola E, Wiesel TN. Increase in retinal vasoactive intestinal polypeptide after eyelid fusion in primates. Proc Natl Acad Sci U S A. 1988;85(1):257–60. pmid:2448769
- 56. Wang Y, Li L, Tang X, Fan H, Song W, Xie J, et al. The role of vasoactive intestinal peptide (VIP) in atropine-related inhibition of the progression of myopia. BMC Ophthalmol. 2024;24(1):41. pmid:38279089
- 57. Mizumoto S, Yamada S. The Specific Role of Dermatan Sulfate as an Instructive Glycosaminoglycan in Tissue Development. Int J Mol Sci. 2022;23(13). pmid:35806490
- 58. Mizumoto S, Yamada S. Histories of Dermatan Sulfate Epimerase and Dermatan 4-O-Sulfotransferase from Discovery of Their Enzymes and Genes to Musculocontractural Ehlers-Danlos Syndrome. Genes (Basel). 2023;14(2). pmid:36833436
- 59. Sun Y, Kuek V, Qiu H, Tickner J, Chen L, Wang H, et al. The emerging role of NPNT in tissue injury repair and bone homeostasis. J Cell Physiol. 2018;233(3):1887–94. pmid:28513838
- 60. Akwii RG, Sajib MS, Zahra FT, Mikelis CM. Role of Angiopoietin-2 in Vascular Physiology and Pathophysiology. Cells. 2019;8(5). pmid:31108880
- 61. Singh RK, Kumar S, Kumar S, Shukla A, Kumar N, Patel AK, et al. Potential implications of protein kinase Cα in pathophysiological conditions and therapeutic interventions. Life Sci. 2023;330:121999.
- 62. Qin Y, Liu T, Zhang Z, Xing S, Gong L, Ni Y. Scleral remodeling in early adulthood: the role of FGF-2. Sci Rep. 2023;13(1):20779. pmid:38012225
- 63. Stone RA, Lin T, Laties AM. Muscarinic antagonist effects on experimental chick myopia. Exp Eye Res. 1991;52:755–8. pmid:1855549
- 64. Arumugam B, McBrien NA. Muscarinic antagonist control of myopia: evidence for M4 and M1 receptor-based pathways in the inhibition of experimentally-induced axial myopia in the tree shrew. Invest Ophthalmol Vis Sci. 2012;53(9):5827–37. pmid:22836762
- 65. Genaro K, Luo ZD. Pathophysiological roles of thrombospondin-4 in disease development. Semin Cell Dev Biol. 2024;155(Pt B):66–73. pmid:37391348
- 66. Gao Q, Chen X, Duan H, Wang Z, Feng J, Yang D, et al. FXYD6: a novel therapeutic target toward hepatocellular carcinoma. Protein Cell. 2014;5(7):532–43. pmid:24715268
- 67. Crewther DP. The role of photoreceptors in the control of refractive state. Prog Retin Eye Res. 2000;19(4):421–57. pmid:10785617
- 68. Crewther SG, Liang H, Junghans BM, Crewther DP. Ionic control of ocular growth and refractive change. Proc Natl Acad Sci U S A. 2006;103(42):15663–8. pmid:17023537
- 69. Fischer AS, Sigal M. The Role of Wnt and R-spondin in the Stomach During Health and Disease. Biomedicines. 2019;7(2). pmid:31248166
- 70. Liu Z, Xiu Y, Qiu F, Zhu Z, Zong R, Zhou X, et al. Canonical Wnt Signaling Drives Myopia Development and Can Be Pharmacologically Modulated. Invest Ophthalmol Vis Sci. 2021;62(9):21. pmid:34259818
- 71. Wu Y, Chen P, Sun L, Yuan S, Cheng Z, Lu L, et al. Sepiapterin reductase: Characteristics and role in diseases. J Cell Mol Med. 2020;24(17):9495–506. pmid:32734666
- 72. Francisco BM, Salvador M, Amparo N. Oxidative stress in myopia. Oxid Med Cell Longev. 2015;2015:750637. pmid:25922643
- 73. George A, Schmid KL, Pow DV. Retinal serotonin, eye growth and myopia development in chick. Exp Eye Res. 2005;81(5):616–25. pmid:15949800
- 74. Zhou X, Pardue MT, Iuvone PM, Qu J. Dopamine signaling and myopia development: What are the key challenges. Prog Retin Eye Res. 2017;61:60–71. pmid:28602573
- 75. Gu X, Nardone C, Kamitaki N, Mao A, Elledge SJ, Greenberg ME. The midnolin-proteasome pathway catches proteins for ubiquitination-independent degradation. Science. 2023;381(6660):eadh5021. pmid:37616343
- 76. Li D, Wang TW, Aratani S, Omori S, Tamatani M, Johmura Y, et al. Transcriptomic characterization of Lonrf1 at the single-cell level under pathophysiological conditions. J Biochem. 2023;173(6):459–69. pmid:36888978
- 77. Ghafouri-Fard S, Khoshbakht T, Taheri M, Shojaei S. A Review on the Role of SPRY4-IT1 in the Carcinogenesis. Front Oncol. 2021;11:779483. pmid:35096580
- 78. Gupta A, Towers C, Willenbrock F, Brant R, Hodgson DR, Sharpe A, et al. Dual-specificity protein phosphatase DUSP4 regulates response to MEK inhibition in BRAF wild-type melanoma. Br J Cancer. 2020;122(4):506–16. pmid:31839677
- 79. Ratsada P, Hijiya N, Hidano S, Tsukamoto Y, Nakada C, Uchida T, et al. DUSP4 is involved in the enhanced proliferation and survival of DUSP4-overexpressing cancer cells. Biochem Biophys Res Commun. 2020;528(3):586–93. pmid:32505357
- 80. Morrison DK. MAP kinase pathways. Cold Spring Harb Perspect Biol. 2012;4(11). pmid:23125017
- 81. Posey KL, Coustry F, Hecht JT. Cartilage oligomeric matrix protein: COMPopathies and beyond. Matrix Biol. 2018;71–72:161–73. pmid:29530484
- 82. van der Sande E, Haarman AEG, Quint WH, Tadema KCD, Meester-Smoor MA, Kamermans M, et al. The Role of GJD2(Cx36) in Refractive Error Development. Invest Ophthalmol Vis Sci. 2022;63(3):5. pmid:35262731
- 83. Kothmann WW, Massey SC, O’Brien J. Dopamine-stimulated dephosphorylation of connexin 36 mediates AII amacrine cell uncoupling. J Neurosci. 2009;29(47):14903–11. pmid:19940186
- 84. Li H, Zhang Z, Blackburn MR, Wang SW, Ribelayga CP, O’Brien J. Adenosine and dopamine receptors coregulate photoreceptor coupling via gap junction phosphorylation in mouse retina. J Neurosci. 2013;33(7):3135–50. pmid:23407968
- 85. Wu Q, Xu WD, Huang AF. Role of angiopoietin-2 in inflammatory autoimmune diseases: A comprehensive review. Int Immunopharmacol. 2020;80:106223. pmid:31991374
- 86. Joussen AM, Ricci F, Paris LP, Korn C, Quezada-Ruiz C, Zarbin M. Angiopoietin/Tie2 signalling and its role in retinal and choroidal vascular diseases: a review of preclinical data. Eye (Lond). 2021;35(5):1305–16. pmid:33564135
- 87. Vasconcelos AR, Dos Santos NB, Scavone C, Munhoz CD. Nrf2/ARE Pathway Modulation by Dietary Energy Regulation in Neurological Disorders. Front Pharmacol. 2019;10:33. pmid:30778297
- 88. Gisbert S, Wahl S, Schaeffel F. L-opsin expression in chickens is similarly reduced with diffusers and negative lenses. Vision Res. 2023;210:108272. pmid:37269575
- 89. Hendriks M, Verhoeven VJM, Buitendijk GHS, Polling JR, Meester-Smoor MA, Hofman A, et al. Development of Refractive Errors-What Can We Learn From Inherited Retinal Dystrophies? Am J Ophthalmol. 2017;182:81–9. pmid:28751151
- 90. Giummarra L, Crewther SG, Riddell N, Murphy MJ, Crewther DP. Pathway analysis identifies altered mitochondrial metabolism, neurotransmission, structural pathways and complement cascade in retina/RPE/ choroid in chick model of form-deprivation myopia. PeerJ. 2018;6:e5048. pmid:29967729
- 91. Vocale LG, Crewther S, Riddell N, Hall NE, Murphy M, Crewther D. RNA-seq and GSEA identifies suppression of ligand-gated chloride efflux channels as the major gene pathway contributing to form deprivation myopia. Sci Rep. 2021;11(1):5280. pmid:33674625
- 92. Tkatchenko TV, Troilo D, Benavente-Perez A, Tkatchenko AV. Gene expression in response to optical defocus of opposite signs reveals bidirectional mechanism of visually guided eye growth. PLoS Biol. 2018;16(10):e2006021. pmid:30300342
- 93. Meyer C, Mueller MF, Duncker GI, Meyer HJ. Experimental animal myopia models are applicable to human juvenile-onset myopia. Surv Ophthalmol. 1999;44 Suppl 1:S93–102. pmid:10548121
- 94. Sahoo AR, Buck M. Structural and Functional Insights into the Transmembrane Domain Association of Eph Receptors. Int J Mol Sci. 2021;22(16). pmid:34445298
- 95. Summers JA, Jones KL. Single Cell Transcriptomics Identifies Distinct Choroid Cell Populations Involved in Visually Guided Eye Growth. Front Ophthalmol (Lausanne). 2023;3.
- 96.
Laties AM, Stone RA. Chap. 35, Ametropia in retinal disorders. In: Anderson RE, Hollyfield JG, LaVail MM, editors. Retinal Degenerations. Boca Raton: CRC Press; 1991. p. 383–90.
- 97. Harb EN, Wildsoet CF. Origins of Refractive Errors: Environmental and Genetic Factors. Annu Rev Vis Sci. 2019;5:47–72. pmid:31525141
- 98. Ayaki M, Torii H, Tsubota K, Negishi K. Decreased sleep quality in high myopia children. Sci Rep. 2016;6:33902. pmid:27650408
- 99. Jee D, Morgan IG, Kim EC. Inverse relationship between sleep duration and myopia. Acta Ophthalmol. 2016;94(3):e204–10. pmid:26031352
- 100. Li R, Chen Y, Zhao A, Huang L, Long Z, Kang W, et al. Relationships between Sleep Duration, Timing, Consistency, and Chronotype with Myopia among School-Aged Children. J Ophthalmol. 2022;2022:7071801. pmid:35903175
- 101. Dong XX, Xie JY, Li DL, Dong Y, Zhang XF, Lanca C, et al. Association of sleep traits with myopia in children and adolescents: A meta-analysis and Mendelian randomization study. Prev Med. 2024;180:107893. pmid:38342383
- 102. Jin E, Lee CE, Li H, Tham YC, Chen DZ. Association between sleep and myopia in children and adolescents: a systematic review and meta-analysis. Graefes Arch Clin Exp Ophthalmol. 2023. pmid:38091060
- 103. Lee SSY, Nilagiri VK, Mackey DA. Sleep and eye disease: A review. Clin Exp Ophthalmol. 2022;50(3):334–44. pmid:35263016
- 104. Li M, Tan CS, Xu L, Foo LL, Yap F, Sun CH, et al. Sleep Patterns and Myopia Among School-Aged Children in Singapore. Front Public Health. 2022;10:828298. pmid:35400064
- 105. Stafford-Bell N, McVeigh J, Lingham G, Straker L, Eastwood PR, Yazar S, et al. Associations of 12-year sleep behaviour trajectories from childhood to adolescence with myopia and ocular biometry during young adulthood. Ophthalmic Physiol Opt. 2022;42(1):19–27. pmid:34676908
- 106. Huang L, Chen X, Lin J, Fan X, Chen T, Yu Y, et al. Association between sleep duration and myopia among Chinese children during the COVID-19 pandemic: A cross-sectional study. Front Public Health. 2022;10:1015138. pmid:36699911
- 107. Liu XN, Naduvilath TJ, Wang J, Xiong S, He X, Xu X, et al. Sleeping late is a risk factor for myopia development amongst school-aged children in China. Sci Rep. 2020;10(1):17194. pmid:33057123
- 108. Ostrin LA, Read SA, Vincent SJ, Collins MJ. Sleep in Myopic and Non-Myopic Children. Transl Vis Sci Technol. 2020;9(9):22. pmid:32879778
- 109. Yu D, Wang L, Zhou X, Liu L, Wu S, Tang Q, et al. Sleep Quality is Associated with Axial Length Elongation in Myopic Children Receiving Orthokeratology: A Retrospective Study. Nat Sci Sleep. 2023;15:993–1001. pmid:38050564
- 110. Archer SN, Schmidt C, Vandewalle G, Dijk DJ. Phenotyping of PER3 variants reveals widespread effects on circadian preference, sleep regulation, and health. Sleep Med Rev. 2018;40:109–26. pmid:29248294
- 111. Dhakal R, Shah R, Huntjens B, Verkicharla PK, Lawrenson JG. Time spent outdoors as an intervention for myopia prevention and control in children: an overview of systematic reviews. Ophthalmic Physiol Opt. 2022;42(3):545–58. pmid:35072278
- 112. Chakraborty R, Seby C, Scott H, Tang V, Kemps E, Anstice N, et al. Delayed melatonin circadian timing, lower melatonin output and sleep disruptions in myopic, or short-sighted, children. Sleep. 2023.
- 113. Fulton JM, Flanagan SC, Sittlington JJ, Cobice D, Dobbin S, McCullough SJ, et al. A Cross-Sectional Study of Myopia and Morning Melatonin Status in Northern Irish Adolescent Children. J Ophthalmol. 2023;2023:7961623. pmid:37946723
- 114. Hussain A, Gopalakrishnan A, Scott H, Seby C, Tang V, Ostrin L, et al. Associations between systemic melatonin and human myopia: A systematic review. Ophthalmic Physiol Opt. 2023;43(6):1478–90. pmid:37568264
- 115. Biswas S, El Kareh A, Qureshi M, Lee DMX, Sun CH, Lam JSH, et al. The influence of the environment and lifestyle on myopia. J Physiol Anthropol. 2024;43(1):7. pmid:38297353
- 116. Harrington S. Shifting landscapes: The environmental impact of urbanisation on childhood myopia, obesity and diabetes. Ophthalmic Physiol Opt. 2024;44(2):237–40. pmid:38078538
- 117. Hobday R. Myopia and daylight in schools: a neglected aspect of public health? Perspect Pub Health. 2016;136:50–5.
- 118. Schaeffel F. Myopia ‐ what is old and what is new? Optom Vis Sci. 2016;93:1022–30. pmid:27429417
- 119. Falchi F, Bará S. Light pollution is skyrocketing. Science. 2023;379(6629):234–5. pmid:36656943
- 120. Wang LB, Gong YC, Fang QL, Cui XX, Dharmage SC, Jalaludin B, et al. Association Between Exposure to Outdoor Artificial Light at Night and Sleep Disorders Among Children in China. JAMA Netw Open. 2022;5(5):e2213247. pmid:35594042
- 121. Quinn GE, Shin CH, Maguire MG, Stone RA. Myopia and ambient lighting at night. Nature. 1999;399:113–4. pmid:10335839
- 122. Harrington S , O’Dwyer V. The association between time spent on screens and reading with myopia, premyopia and ocular biometric and anthropometric measures in 6- to 7-year-old schoolchildren in Ireland. Ophthalmic Physiol Opt. 2023. pmid:36843144
- 123. Liu XN, Naduvilath TJ, Sankaridurg PR. Myopia and sleep in children-a systematic review. Sleep. 2023;46(11). pmid:37381700
- 124. Zhu Z, Chen Y, Tan Z, Xiong R, McGuinness MB, Müller A. Interventions recommended for myopia prevention and control among children and adolescents in China: a systematic review. Br J Ophthalmol. 2023;107(2):160–6. pmid:34844916
- 125. Sánchez-Tena M, Ballesteros-Sánchez A, Martinez-Perez C, Alvarez-Peregrina C, De-Hita-Cantalejo C, Sánchez-González MC, et al. Assessing the rebound phenomenon in different myopia control treatments: A systematic review. Ophthalmic Physiol Opt. 2024;44(2):270–9. pmid:38193312